Fix parsing of GLYCAM topology files

[lohedges]

This PR fixes several bugs affecting systems parameterised with the GLYCAM force field (Glycam-06), which uses full 1-4 non-bonded interactions (SCEE=1.0, SCNB=1.0) rather than the standard AMBER scaling (SCEE=1.2, SCNB=2.0). GLYCAM topologies from CHARMM-GUI represent these interactions using [pairs] funct=2 with explicit parameters, rather than funct=1 which applies the global fudgeLJ/fudgeQQ scaling from [defaults].

Bugs fixed:

GROMACS read (grotop.cpp): [pairs] entries were listed in missed_tags and completely ignored on read. All 1-4 pairs were assigned the global fudgeQQ/fudgeLJ scaling factors, so GLYCAM's full 1-4 interactions were silently replaced with standard AMBER scaling. Fixed by parsing funct=2 entries and storing them as explicit pair overrides applied after the gen-pairs CLJNBPairs is constructed.

GROMACS write (grotop.cpp): CLJScaleFactor(1.0, 1.0) pairs fell through the write_pairs logic without being written. Since nrexcl=3 excludes all 1-4 pairs by default, omitting them from [pairs] gives zero 1-4 interaction. Fixed by adding a funct=2 write branch that computes combined LJ parameters from atom types using the topology's combining rule.

AMBER write (amberparams.cpp): Two conditions incorrectly skipped CLJScaleFactor(1.0, 1.0) pairs, causing them to be silently dropped and written as SCEE=0/SCNB=0 in the prm7 file. Fixed by correcting both conditions to exclude only fully-excluded (0.0, 0.0) pairs.

GROMACS read (gro87.cpp, grotop.cpp): Residue numbering that restarts at 1 (common in protein+glycan complexes) caused duplicate ResNum exceptions or multi-minute hangs due to O(n²) conflict resolution. Fixed by tracking residue boundaries via name+number rather than number alone, with O(1) unique numbering.

BioSimSpace merge (_merge.py, biosimspace.{h,cpp}): The merged molecule intrascale properties were constructed from global forcefield scale factors rather than the actual per-pair values stored in the molecule's intrascale property. For GLYCAM this caused an energy error in the merged end states. Fixed via a new C++ function SireIO::mergeIntrascale that directly copies per-pair scale factors into the merged molecule's atom index space using a two-pass approach, without relying on CLJNBPairs::merge which reconstructs pairs from global ff parameters.

• I confirm that I have merged the latest version of devel into this branch before issuing this pull request (e.g. by running git pull origin devel): [y]
• I confirm that I have added a changelog entry to the changelog (we will add a link to this PR as part of the review): [y]
• I confirm that I have permission to release this code under the GPL3 license: [y]

[lohedges]

It looks like the reindexing fix has broken one of my tests. It also appears that I need to make changes in AmberParams to account for the GLYCAM updates.